Connectivity drivers

[SeppoTakalo]

Description

Import connectivity drivers into Mbed OS.

• ESP8266 (Wifi)
• Atmel AT86RF233/AT86RF212 (IEEE 802.15.4)
• MCR20A (IEEE 802.15.4)

After this change, any of those driver can be selected as a default driver. This allows building connectivity examples for applications without connectivity.
This allows us to migrate away from easy-connect dependency.

Documentation how default network interfaces work, please see:

• Add information about default network interface mbed-os-5-docs#660

This PR depends on #7778

Pull request type

[ ] Fix
[ ] Refactor
[ ] Target update
[X] Functionality change
[ ] Breaking change

[SeppoTakalo]

@kjbracey-arm please review

[jeromecoutant]

Hi Seppo

Can I ask why wifi-ism43362 is not imported as well ?

Thx

[SeppoTakalo]

Hi @jeromecoutant for the first round I'm importing only the ones that we are actively maintaining and testing in our CI.

This is part of the change where I submitted PR earlier into most of the connectivity drivers, ISM as well in ARMmbed/wifi-ism43362#27
Long term goal is to slowly migrate away from easy-connect repository.

Now going forward, applications can be made portable by just choosing which one to provide the default interface.

Driver being outside of Mbed OS tree does not make any difference, once the configuration is in place.

[SeppoTakalo]

@cmonr What is the justification for "Needs work" label?

[0xc0170]

As this is adding components, doesn't this depend on components tools changes or is it good as it is?

[SeppoTakalo]

@0xc0170 This does not depend on the component support, even that the folder name is the same.

These drivers have been tested not to affect RAM or ROM usage, as long as they are unused. So there is no need to exclude from the build.

[SeppoTakalo]

Rebased on top of master.
MCR20A driver changed so that it is enabled if build against KW24D target.

[cmonr]

@SeppoTakalo I would normally say to check the failure logs, but 1) the job artifacts have been garbage collected by Jenkins due to the history length, and 2) I'm not sure how to retrigger a build.

@ARMmbed/mbed-os-test Some help?

[AnttiKauppila]

This is dependent on #7778 which should go in first

[SeppoTakalo]

Client test is known to fail, because it is testing with previous release.
Once this goes in, there are fixes to client in the internal repositories. PRs already submitted.

[cmonr]

@SeppoTakalo @0xc0170 Keep in mind that @bulislaw is on vacation for the rest of the month.

@cmonr What is the justification for "Needs work" label?

I think it was because the cloud_client_smoke_test check was failing. Still talking with @ARMmbed/mbed-os-test to see how to restart the test.

[cmonr]

Now waiting on @kjbracey-arm and the predeeding PR.

[SeppoTakalo]

Rebased on top of master.

[mbed-ci]

Build : FAILURE

Build number : 2981
Build artifacts/logs : http://mbed-os.s3-website-eu-west-1.amazonaws.com/?prefix=builds/7814/

[cmonr]

I derp'd and used origin/master instead of upstream/master. Doing the rebase again.

[cmonr]

/morph build

[mbed-ci]

Build : SUCCESS

Build number : 2982
Build artifacts/logs : http://mbed-os.s3-website-eu-west-1.amazonaws.com/?prefix=builds/7814/

Triggering tests

/morph test
/morph export-build
/morph mbed2-build

[mbed-ci]

Test : SUCCESS

Build number : 2739
Test logs :http://mbed-os-logs.s3-website-us-west-1.amazonaws.com/?prefix=logs/7814/2739

[mbed-ci]

Exporter Build : SUCCESS

Build number : 2590
Build artifacts/logs : http://mbed-os.s3-website-eu-west-1.amazonaws.com/?prefix=builds/exporter/7814/

[pan-]

In which world this doesn't get optimized out (ref) ?

[SeppoTakalo]

It does, I just needed a placeholder to get it building with ARMC6.

Other compilers already have assembly version provided, but none of those builded OK with ARMC6

[geky]

Should this be in a COMPONENT_ESP8266 directory?